Introduction

Cork world production drop from 380,000 t in 1999to 299,300 t in 2008 (Aronson et al., 2009). Thoughbearing a traditional management of the cork oaks stands,the part of Algeria in the cork market (414,000 ha and15,000 t) currently shows a rate of 5%, which impliesa low production (e.g., Italy with 92,000 ha, produced17,000 t) (Pereira et al., 2008). Meanwhile, due to itsnature and the heterogeneity of the factors affectingits production, cork is a product with a high variability,and so the concept of quality is difficult to define (Carras-quinho, 1987).

Through its formation process, changes occur in thecork due to environmental perturbations that affect itsgrowth, color, texture, density and frequency of its po-rosity (Molinas and Oliva, 1990). Among these altera-tions is necessary to quote the decay. This is a clear,persistent and progressive decline of the trees extendedbeyond a region or forest. This phenomenon is charac-terized by the appearance of abnormal leaves (yello-wish, premature fall, necrosis, small size, clustered),the death of the trunk from the branch tips toward thestem, and the death of rootlets and reduction in radialgrowth (Auclair, 1993). The decay is related to severalfactors including periods of exceptional drought, ad-verse soil conditions, absent or unsuitable silviculturalpractices, fires and plagues of insects or other parasites(Bakry and Abourouh, 1996).

Short communication. Effect of the health status on the corkproduction characteristics of Western Algeria cork oak stands

B. Dahane1, R. T. Bouhraoua1, J. L. Garcia de Ceca2* and J. R. Gonzalez-Adrados2

1 Départament de Foresterie. Faculté des Sciences. BP 119. Université de Tlemcen. Argelia2 INIA. Centro e Investigación Forestal. Departamento de Productos Forestales.

Ctra. A Coruña, km 7,5. 28040 Madrid, Spain

Abstract

Aim of study: To analyze the effect of health status on cork production, analyzing if this influence is uniform or isaffected by site conditions.

Area of study: Two Western Algerian cork tree forests have been studied: M’Sila located in the coastal plains undersemiarid climate, and Zarieffet, located in the mountainous interior under sub-humid climate.

Material and methods: 40 trees were selected in each forest and classified according to their health status as healthy,weakened, or decaying. A sample of cork from each tree has been obtained to measure the key variables related to corkproduction. A two-way ANOVA was performed considering two factors: site and health status.

Main results: Quercus suber L. productivity is affected by the vitality of trees in the same way in both sources, showingvalues between 5.96 ± 7.1 kg · m–2 (coast, weakened trees) and 8.13 ± 0.45 kg · m–2 (mountain, healthy trees). The healthstatus also affects the number and area of pores, especially in the cork oak groves of the coast, where the coefficient ofporosity ranges from 3.79 ± 0.84% (healthy trees) to 8.11 ± 1.91% (decaying trees). The variables where the site has presenteda stronger effect are those related to the amount of cork produced by the phellogen (density —kg · l–1, p < 0.000— andproductivity —kg · m–2, p = 0.001—), and pore density (1 · cm–2, p = 0.001). Scrap thickness (mm) and porosity (%) showa smaller effect although still representative (p = 0.041 and 0.038 respectively). Porosity and pore density show interactionsite*health status. They all have higher values in the mountain (Zarieffet) than in the coast (M’Sila). The effect of treevitality on the formation of pores in the cork oak phellogen is lower in the mountain than in the coast. No significant effectswere found for any of the two factors neither on the annual growth rate nor on the thickness of the cork.

Research highlights: Results lead to the conclusion that the effect of health status on traumatic phellogen formationand activity is clear but not uniform. Further studies are necessary for a deeper understanding of the effect of stresssituations on pore formation and characteristics.

Key words: decay; density; mean annual growth; porosity; productivity.

* Corresponding author: ceca@inia.esReceived: 06-03-12. Accepted: 03-02-13.

Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA) Forest Systems 2013 22(1), 138-146 Available online at www.inia.es/forestsystems ISSN: 2171-5068http://dx.doi.org/10.5424/fs/2013221-02866 eISSN: 2171-9845

The problems of health degradation of forests (de-ciduous and conifers) are very old, dating from the latenineteenth and early twentieth century in many coun-tries, especially in Europe, but also in North America.But the emergence of the most dramatic decline hasbeen in the early years of the eighties (Delatour, 1983;Bonneau and Landmann, 1988). In Algeria, the healthproblems of cork oak forests date back to early lastcentury, especially in the middle and even towards theWest of the country, appearing the decay in young trees(under 15 years of age) as in mature trees. At mid-twentiethcentury, many populations already presented severesymptoms of weakness (Boudy, 1952).

This work has two main objectives:1. To study how the cork production variables

(growth rate, density, productivity, porosity, pore den-sity, size and thickness of the scrap) are influenced bythe sanitary state of the trees (healthy, weak, decadent).

2. To study whether the geographic origin canaffect the behavior of these variables.

To do this there have been selected two represen-tative forest of western Algerian cork area: in the litto-ral zone, M’Sila forest in Boutlelis municipality; inthe mountain area, Zarieffet forest in Tlemcen munici-pality (Fig. 1, Table 1). In M’Sila, the phenomenon of decay of the trees is f irst mentioned in 1891(Bouhraoua, 2003). Other waves of decay were re-corded then between 1913 and 1926. From 1975 (dateof first use of the cork after the Independence of Alge-

ria), the decay of the forests called attention to the localadministration, but the real work to diagnose and iden-tify possible causes of the problem did not begin untilthe late 80. New field studies were organized between1983 and 1995 (Lanier et al., 1986; Aici et al., 1994). Du-ring this period there was a net increase of decay, evenlydistributed throughout the forest in the form of spots. Thedeteriorating health of Zarieffet forests began in the earlytwentieth century and persists to date (AEFCO, 1969),having observed a significant tree mortality in the late80’s. Data are available from the monitoring network forthese forests, collected annually between 1999 and 2008.Inventories conducted for the two forests show signifi-cant increases in the percentage of trees with leaf defi-cit greater than 25% between 2000 and 2006, reachingvalues above 60% in M’Sila and 45% in Zarieffet.

The health index Is (Bouhraoua, 2003), was used torepresent the health status of the forests. This index allowsexpressing directly the general state of the populationfrom a set of trees taken individually.

It is calculated as follows:

where:ni = number of trees of defoliation type i.Pi = weight of type i (1 if i = 1; 2 if = 2, etc.).N = observed total number of observed trees.

Short communication. Health status effect on the cork production characteristics of Algerian cork oak stands 139

Figure 1. Geographical sample points: Boutlelis (M’Sila) and Tlemcen (Zarieffet).

Tlemcen

Ghazaouet

Sebdou

Marsat BenMhidi

Morocco

Ain Temouchent

Boutlelis

Oran

ALGERIA

Mediterranean Sea

ALGERIA

Tlemcen

The criterion used is based on the defoliation resul-ting in the eventual loss of leaves in the functional partof the crown. It is an indicator widely used in Euro-pe to assess the health status of forest trees (DSF,1991). It was taken in summer by visual examinationof the ramif ications of the tree, assigning to each tree one of the following classes that correspond to ahealth category (Bouhraoua and Villemant, 2002)(Fig. 2):

— 1 (< 25%, healthy trees).— 2 (25-60%, weak trees).— 3 (> 60%, decadent trees).The health assessment was based on annual monito-

ring from 1999 to 2008. The average of all visual obser-vations gives the value of the final health status of eachtree (Bouhraoua, 2003; Dahane, 2006).

Calculated for each forest (Table 1), on the coast Isreveals a more serious situation which can be describedas «self-evident weakening» in the groves of M’Sila(2.1 < Is < 2.5). On the contrary, the situation on the moun-tain is more satisfying; being the cork oak grove of Zarieffeton the edge of the qualification of «healthy» (Is < 1.5).

The progressive deterioration of the two popula-tions, both showing a similar trend of degradation andimpoverishment, implies the destruction of cork wor-king capital. Moreover, the cork annual growth is di-rectly influenced by the decay (Cambini, 1971). Thisimplies a reduction in both cork thickness and producti-vity of around 42% and 50% respectively. No studies werefound concerning the influence of decay on other corkproperties (porosity, density...) that were directly rela-ted to industrial quality and economic value of the product.

140 B. Dahane et al. / Forest Systems (2013) 22(1), 138-146

Table 1. Characteristics of sampling points

Characteristics M’Sila Zarieffet

Mean altitude (m) 380 1,100Topography Plateau Upper basinSlope (%) 3 30Origin of populations Natural NaturalStructure of populations Irregular IrregularTreatment High Forest System High Forest SystemNature of the population Mixed PureMean density (tree · ha–1) 160 90Forest fires (last occurrence) — 1994-2007Average height of trees (m) 8.35 ± 0.39 8.40 ± 1.07Average circumference of trees (cm) 99.09 ± 6.6 128.11 ± 35.08Average height debarking (m) 2.44 ± 0.35 2 .26 ± 0.35Mean coefficient of debarking 2.47 ± 0.18 1.76 ± 0.33Health status (Is) 2.30 1.49

Figure 2. Defoliation type 1, 2, and 3 (healthy, weak, and decaying trees).

Material and methods

There was obtained a cork sample from two forestsin Northwest Algeria (Table 1). In the littoral zone, theM’Sila forest (1,087 ha) was selected, with a semiaridclimate characterized by mild winters and hot, humidsummers. In the mountainous area, the Zarieffet forest(960 ha) was selected, with an inferior subhumid cli-mate characterized by cold and rainy winters and hotdry summers. Both formations are integrated into themonitoring network health status of cork oak forestsin Western Algeria, established since 1999.

Table 2 (Average monthly temperatures and annualT (°C) in both forests) shows that average annual tem-peratures in the littoral are of the order de 18.1°C against15.9°C Zarieffet. January is usually the coldest monthfor both forests (12°C and 7°C M’Sila to Zarieffet).The warmest month is denoted by cons in August with24.8 °C for the first zone and 25.6°C for the second.

Table 3 (Mean monthly precipitation (mm) duringthe two reference periods) shows that the mean monthlyrainfall varies from one period to another. In the forestof M’Sila, the rainfall is more intense during the monthsof November and December (> 60 mm). Other monthsshow less pronounced rainfall ranges mostly betweenMay and September, indicating a signif icant waterdeficit.

In the forest of Zarieffet, the rainy season starts inpart from September and reached its peak during themonths from December to March (>60 mm). The monthsfrom June to August are still dry.

According to Gourinard (1958), the M’Sila forestcovers an old Callabrian surface, still covered locallywith marine deposits or correspondent dune. The cork

oak is based on a very deep soil (> 2 m). They are cove-red with sand formations, coarse-textured grain size,fairly uniform from one station to another and fromone horizon to another. The upper horizons, dry andmostly siliceous, contain a very small fraction of clay(< 3%) which signif icantly increases to over 60 cmdepth (Bouhraoua, 2003).

Zarieffet forest at Tlemcen is based on geologicalstrata that were formed mainly on an Upper Jurassicmassif from a tertiary fold (Benest, 1985). The bedrockis composed mainly of Sequanian limestone, Kimmeri-dgian and Portlandian dolomites, and scattered out-crops. The soils are generally crossed by rock surfacesand rocky outcrops, being essentially superficial, atleast 30 cm deep. These soils have a very heteroge-neous texture from one station to another, ranging fromsandy loam to sandy clay. Clay formations do not appearin this forest but in some places in the upper horizonsof sandy clay soils. Nevertheless, they are often presentbeyond 80 cm depth (Kaïd Slimane, 2000).

Four circular plots of 20 m radius were marked ineach forest, ranging the number of trees per plot bet-ween 25 and 30. Ten trees were sampled at random foreach plot, extracting from each tree at 1.30 m high andfacing the center of the plot a rectangular sample ofcork with dimensions of 15 × 15cm (225 cm2). The ro-tation age to obtain the sample was of 9 years. In total,40 samples were obtained from each forest.

All plots belonged to a surveillance network dedi-cated to control the health status of the trees. In eachtree, the health status was assessed in the summer byvisual examination of the crown. Each tree was assig-ned to one of the following health categories (DSF,1991; Bouhraoua and Villemant, 2002) (Fig. 2):

Short communication. Health status effect on the cork production characteristics of Algerian cork oak stands 141

Table 2. Monthly average precipitation (mm) for two periods of time

Site Period J F M A M J Jl A S O N D Total

M’Sila 1913-34 71.3 59.4 48.6 40.0 31.3 10.8 1.08 3.2 20.5 44.3 82.1 86.4 5011971-01 61.5 50.0 49.2 31.9 25.7 5.2 0.2 1.5 11.5 38.0 63.1 61.9 400

Zarieffet 1913-38 83.2 101 93.2 72.3 65.5 27.7 2.2 5.6 26.6 57.7 92.2 81.0 710.51961-04 76.0 75.3 103 82.9 62.2 16.7 4.00 3.3 23.2 45.5 63.9 71.3 625.5

Table 3. Average monthly and yearly temperatures (°C) for both forests

Site Period J F M A M J Jl A S O N D Mean

M’Sila 1913-34 11.1 11.6 12.8 14.6 17.0 19.9 22.7 23.5 21.6 18.1 14.6 12.0 16.61971-01 10.5 11.4 12.5 13.8 16.5 19.5 22.8 23.7 20.7 17.5 14.2 11.4 16.2

Zarieffet 1914-38 5.9 7.1 8.8 11.2 15.2 19.6 25.4 25.8 21.1 16 10.1 6.8 14.51975-04 6.7 8.5 10.6 10.9 14.3 20.2 22.5 22.9 20.4 18.9 11.7 8.9 14.7

— Class 1 (defoliation < 25%, healthy trees).— Class 2 (defoliation 25-60%, weak trees).— Class 3 (defoliation > 60%, decadent trees).This classif ication is shorter than the commonly

used in similar works (SPCAN-DGB, 2007), but is bestsuited for our goals.

Being nine years the rotation time, the health assess-ment of each tree of the sample was made from thesurveillance network data files from 1999 to 2008 (yearof sample collection). The average of all visual obser-vations (1, 2, or 3) qualify a tree as healthy (mean ≤1.40), weak (1.40 < mean ≤ 2.40) or declining (mean >2.40). The variables to consider are: Growth (mm · year–1),thickness (mm), scrap thickness (mm), density (kg · l–1),productivity (kg · m–2), porosity (%), and pore density(1 · cm–2).

Cork samples were put in boiling water for 1 houraccording to the practice of industrial preparation ofcork. Then, they were stabilized at room temperature(20°C) till constant weight to remove the water nettedduring the boiling.

Total thickness was determined for each samplefrom belly to back according ISO 1216:1998 standard.The age of the cork was estimated by counting thenumber of full growths (f.g.) plus the addition of twoincomplete growths (i.g.) (e.g., 9 years = 1/2 i.g. + 8f.g. + 1/2 i.g.). The average annual growth was deter-mined disregarding the two incomplete growths. Thescrape thickness was calculated as the average of deter-minations made in 3 equidistant sites. These measure-ments were taken with the aid of a binocular micros-cope fitted with a graduated scale (micrometer) with1/100 mm precision.

To determine the volumetric density (kg · l–1) andproductivity (kg · m–2) of each sample their length andwidth were measured using a flexible rule of 1 mmaccuracy. To calculate the density, the mass was deter-mined with a precision balance of 0.001 g. Thicknessvalue includes the thickness of the scrape.

The calculation of volumetric density responds tothe following empirical formula:

Density (kg · l–1) = M/V

where V (l) = Volume = length * width * thickness

M (kg) = Mass of the sample

Productivity is given by the formula:

Productivity (kg · m–2) = M/S

where S (m2) = length * width

M (kg) = Mass of the simple

The acquisition of the image of the tangential sec-tion of the 80 samples was performed with scanner.The samples were prepared for their study by imageanalysis. The belly of each sample was sanded on theentire section and then cleaned with compressed air.The porosity of the tangential section was studied byimage analysis (González Adrados et al., 2005), usingthe Olympus software cell^D. As the limit of the ima-ges’ resolution, the pores with a surface less than0.07 mm2 were not computed.

The following parameters were determined for eachsample:

1. Coefficient of porosity (%) of the pores totalsurface in the cork total area.

2. Pore density (number of pores · cm–2).After checking the ANOVA conditions, a two-way

ANOVA model was applied to test the effect of healthstatus and origin on the parameters. Two-way ANOVAis a special case of the linear model, being its form:

yijk = µ + α,j + βi, + γij + εijk

where:yijk = is a matrix of studied variable observations (with

row index i, column index j, and repetition index k).µ = is a constant matrix of the overall meanα·j = is a matrix whose columns are the deviations of

each health statusβi· = is a matrix whose rows are the deviations of each

originγij = is a matrix of interactionsεijk = is a matrix of random disturbances

When interaction was detected, one-way ANOVAwas run separately for each origin.

Results

Results from the image analysis were found anoma-lous in four cases, causing non-normality concerns forporosity and pore density. After analyzed, the fourcases were considered as outliers, and image analysisdata were eliminated from the data set. Thus, normalityand homogeneity of variance conditions were consi-dered sound enough to avoid variables transformation.The results for studied variables are shown in Table 4.

To date, there has not been found any work pu-blished about the topic for cork characterization inAlgeria. On the whole, data shown in Table 4 are com-parable with those published for, Portuguese (Pereira,2007) and Spanish (González-Adrados et al., 1993)

142 B. Dahane et al. / Forest Systems (2013) 22(1), 138-146

cork regions. It has to be highlighted the fact that, accor-ding to Pereira (2007) almost all of the samples belongsto the thin commercial class (thickness less than 27 mm).

The analysis of variance can better explain the effectof origin and health status on the variables taken intoaccount. As an example, Table 5 shows the results forthe coeff icient of porosity. Table 6 summarizes theresults of the other variables. Fig. 3 illustrates the dis-tribution of the data grouped by origin and health status.

The effect of health status is shown in variablesrelated to the presence of pores (porosity, pore density),which are also affected by the origin. Both variablespresent a significant interaction Origin * Health Status(p < 0.01). Fig. 3 and one-way ANOVA help to clarifythe meaning of this interaction: for both variables theeffect of health status is very clear in M’Sila (p < 0.001for porosity and p < 0.05 for pore density in one-way

ANOVA) and does not appear in Zarieffet, where theporosity has very few variations. In M’Sila, however,shifting into a worse state of health always imply anincrease in the average values of porosity, both in termsof percentage and density (Fig. 3, Table 6). Pooling thedata from both areas (Table 6) confirms the effect ofhealth status (p < 0.05) being the coefficient of porosityof healthy trees (3.87 ± 0.56%) lower than the decadent(5.69 ± 1,37%). Similar trend occurs with pore densityvalues (5.91 ± 0.74 vs. 7.89 ± 1.55 pore · cm–2), thoughone-way ANOVA results are not significant.

Focusing on the remaining variables studied, densityand thickness of the scrape are clearly affected by thegeographical origin of the samples, but not by thehealth status. The values in Table 4 show how the den-sity of the cork produced in the mountain area (0.38 ±0.02 kg · l–1) is significantly higher than in the littoral

Short communication. Health status effect on the cork production characteristics of Algerian cork oak stands 143

Table 4. Description of the variables studied. N is the number of total samples. N2 same as N but excluding the outliers

Health AnnualThickness

ScrapDensity Productivity Porosity Pore

Originstatus

N growth(mm)

thickness(kg · l –1) (kg · l–1)

N2(%) (1 · cm–2)

(mm · año–1) (mm)

M’Sila 1 21 2.39 ± 0.23 21.68 ± 1.96 4.29 ± 0.72 0.29 ± 0.02 6.34 ± 0.60 21 3.79 ± 0.84 5.70 ± 1.282 10 2.26 ± 0.40 20.44 ± 3.56 4.24 ± 1.20 0.29 ± 0.02 5.96 ± 1.07 10 4.34 ± 1.14 7.37 ± 1.833 9 2.28 ± 0.35 20.55 ± 3.22 4.24 ± 1.10 0.30 ± 0.02 6.03 ± 0.87 8 7.35 ± 1,39 10.03 ± 1,6

Total 40 2.34 ± 0.17 21.11 ± 1.51 4.27 ± 0.52 0.29 ± 0.02 6.17 ± 0.04 39 4.66 ± 0,74 7.01 ± 0.01

Zarieffet 1 21 2.27 ± 0.16 20.97 ± 1.48 5.62 ± 0.94 0.39 ± 0.02 8.13 ± 0.45 21 3.96 ± 0.76 6.12 ± 0.942 10 2.16 ± 0.17 19.57 ± 1.64 4.76 ± 1.01 0.31 ± 0.02 7.04 ± 0.58 9 4.55 ± 1,61 4.77 ± 1.043 9 2.04 ± 0.41 18.67 ± 3.72 4.91 ± 1.47 0.37 ± 0.05 6.57 ± 0.72 7 3.79 ± 1,60 5.44 ± 4,03

Total 40 2.19 ± 0.13 20.10 ± 1.21 5.25 ± 0.65 0.38 ± 0.02 7.50 ± 0.39 37 4,07 ± 0,65 5.66 ± 0,64

Total 1 42 2.33 ± 0.12 21.32 ± 1.22 4.96 ± 0.62 0.34 ± 0.02 7.23 ± 0.47 42 3.87 ± 0.56 5.91 ± 0.782 20 2.21 ± 0.22 20.00 ± 1.92 4.51 ± 0.70 0.32 ± 0.02 6.50 ± 0.64 19 4.40 ± 0,94 6.13 ± 1.213 18 2.16 ± 0.27 19.61 ± 2.43 4.58 ± 0.88 0.34 ± 0.04 6.30 ± 0.56 15 5.69 ± 1,37 7.89 ± 1.55

Total 80 2.27 ± 0.11 20.61 ± 0.97 4.76 ± 0.42 0.34 ± 0.02 6.84 ± 0.32 76 4.37 ± 0.49 6.36 ± 0.62

Table 5. ANOVA results for Coefficient of Porosity

Inter-subjects effects test

SourceType III sum

df Mean square F Sig.of squares

Corrected model 84.652a 5 16.930 4.274 0.002Intercept 1,346.137a 1 1,346.137 339.857 0.000Origin 17.626a 1 17.626 4.450 0.038Health status 32.161a 2 16.081 4.060 0.021Origin * health status 42.035a 2 21.017 5.306 0.007Error 277.262a 70 3.961

Total 1,817.727a 76

Corrected total 361.915a 75

Dependent variable: tg CP (%). a R Squared = 0.234 (Adjusted R Squared = 0.179).

zone (0.29 ± 0.02 kg · l–1). This increase can be attribu-ted, at least partially, to the increased thickness of thescrape (5.25 ± 0.65 mm in Zarieffet compared to 4.27 ±0.52 mm in M’Sila), a woody tissue with bigger densitythan cork itself. Though the differences are not signi-ficant, it has to be said in this regard, that the lowestannual growth rate of cork appears in the mountainareas of Zarieffet, 2.19 ± 0.13 mm · year–1 (2.34 ± 0.17mm · year–1 in M’Sila).

Finally, productivity is affected by the origin as wellas the sanitary status, although with less intensity forthe latest. The effect of this last factor is the same inboth areas (no interaction), being more obvious the diffe-rences (Table 4) between healthy and decadent trees inZarieffet (8.13 ± 0.45 vs. 6.57 ± 0.72 kg · m–2) than inM’Sila (6.34 ± 0.60 vs. 5.96 ± 1.07 kg · m–2). Conside-ring all samples from both forests, the results corres-pond directly with those obtained for the density andthickness of the scrape: productivity is significantlyhigher in Zarieffet (7.50 ± 0.39 kg · m–2) than in M’Sila(6.17 ± 0.04 kg · m–2) (p < 0.001). Neither growth northe caliber achieves significant results, so it can be saidthat there are not differences for these variables bet-ween the groups considered.

Discussion

Results reflect the differences between cork productionof forests located on the coast and in mountainous areasof Eastern Algeria. The health status varies from one area

to another, so that each forest must be analyzed separately.On the coast (M’Sila), porosity is more pronounced

in the cork of trees with the worst health condition.This is probably due to the fact that these trees sufferconsiderable pressure from bush competition for waterand mineral reserves (Bouhraoua, 2003). It is also aforest with a health deficit and high tree density, loca-ted in a semiarid climate. In this situation, the declineis affecting the proper functioning of the phellogen,causing an increase in porosity, but not in the alreadylow growth rate (Courtois and Masson, 1999).

In the mountains (Zarieffet), the decay negativelyaffects productivity. These trees, located in subhumidclimate with significant slopes, have an overall betterhealth status than those from the coast. In them, meanage is greater (≥ 120 years) and tree density is lowercompared to those of the coast. Thus, cork shows a lategrowth of predominantly small cells of the summer-autumn. Under these conditions, both the lower envi-ronmental drying as well as the cork oak lower phy-siological activity makes the new phellogen appearing,after the debarking, at a shallower depth in the liber.This causes thinning of the scrap and reduces the annualincreases of cork, which altogether triggers a decreasein productivity.

Conclusions

In the studied area, the effect of health status on thecork production variables varies depending on ecolo-

144 B. Dahane et al. / Forest Systems (2013) 22(1), 138-146

Table 6. Results summary of test between subjects effects of the variables studied (ANOVA)

Variable Origin Health status Origin * health status

Growth F 1.592 0.866 0.134p 0.211 0.425 0.875

Thickness F 1.159 1.174 0.111p 0.285 0.315 0.895

Scrap thickness F 4.333 0.804 0.030p 0.041 0.451 0.970

Density F 25.015 0.585 0.638p 0.000 0.560 0.531

Productivity F 13.166 4.072 1.555p 0.001 0.021 0.218

Porosity F 4.450 4.060 5.306p 0.038 0.021 0.007

Pore density F 13.014 3.131 6.471p 0.001 0.050 0.003

p < 0.001%; p < 0.01%; p < 0.05%.

gical and silvicultural conditions of the stands. Theinland mountainous area, of subhumid climate andconsiderable slope, has higher productivity than thecoast area in terms of surface density (cork kg produ-ced by m2 of debarked tree). Also, this variable is signi-ficantly affected by the loss of vigor with no apprecia-ble effect on the porosity.

By contrast, in coastal areas of semi-arid climateand little or no slope, decay does not seem to affect theamount of cork produced by the tree, but the qualitativeaspects of the material. Specifically, porosity is affec-ted both in terms of number of pores (that differ in thephellogen) as in their extension (coefficient of poro-sity).

The geographical origin signif icantly affects thedensity and productivity, taking higher values in Za-rieffet. The health status adversely affects productivityin mountain forests (Zarieffet) and porosity in coastalareas (M’Sila).

These results suggest the need for a deeperunderstanding of the factors that influence theformation and characteristics of the pores afterdebarking in the traumatic phellogen.

Acknowledgements

The authors wish to thank the collaboration of Ma

Luisa Cáceres Esteban, and Lorenzo Ortiz Buiza (INIA-CIFOR).

References

AEFCO, 1969. Fascicule de gestion de la forêt domanialede Zarieffet. Non paginé.

Aici M, Lafer M, Bahakemi S, 1994. Rapport de missioneffectué du 13/06/94 au 15/06/94 dans les conserva-tions des forêts d’Oran et Sidi Bel Abbes Alger. ANF-INRF: 4.

Aronson J, Pereira JS, Pausas JG (eds), 2009. Cork OakWoodlands on the edge. Ecology, adaptive management,and restoration. Island Press.

Auclair AND, 1993. Extreme climatic fluctuations as a causeof forest dieback in the pacific Rim. Water, Air, and SoilPollution 66(3-4): 207-229.

Bakry M, Abourouh M, 1996. Nouvelles donnees sur ledeperissement du chene-liege (Quercus suber L.9 auMaroc). Ann Rech For Maroc 29: 24-39.

Benest M, 1985. Evolution de la plate-forme de l’Ouestalgérien et du Nord-Est marocain au cours du Jurassiquesupérieur et au début du Crétacé: stratigraphie, milieuxde dépôt et dynamique sédimentaire. Laboratoire Géologi-que Lyon, France: 95: 581 pp.

Bonneau M, Landmann G, 1988. Le dépérissement des forêtsen Europe. La Recherche 205(19): 1542-1556.

Boudy P, 1952. Guide du forestier en Afrique du Nord. Paris.Bouhraoua R, Villemant C, 2002. Situation sanitaire de

quelques suberaies de l’oust algérien: impact des xylo-phages. IOBC/WPRS Bull 25(5): 85-92.

Bouhraoua RT, 2003. Situation sanitaire de quelques forêtsde chêne-liège de l’ouest algérien. Etude particulière desproblèmes posés par les insectes. Dept Forest Fac Sci UnivTlemcen. 267 pp.

Cambini A, 1971. Valutazione dei danni causati dagli insettidefogliatori alla querci da sughero. 1°Convegno Regionaldel Sughero, Tempio Pausania.

Carrasquinho MI, 1987. A qualidade da cortiça. Boletim doInstituto dos Productos Florestais-Cortiqa (583): 17-18.

Short communication. Health status effect on the cork production characteristics of Algerian cork oak stands 145

Figure 3. Results obtained in three of the variables studied.

Den

sity

(kg

·l–1

)Pr

oduc

tivity

(kg

·l–1

)Po

rosi

ty (%

)

1 2 3 1 2 3M’Sila Zarieffet

1 2 3 1 2 3M’Sila Zarieffet

1 2 3 1 2 3M’Sila Zarieffet

0.6

0.5

0.4

0.3

0.2

0.1

11

10

9

8

7

6

5

4

12

10

8

6

4

2

0

Courtois M, Masson P, 1999. Relation entre la qualité duliège, sa composition minérale et la composition minéraledes feuilles du chêne-liège (Quercus suber L.). Ann ForSci 56: 521-527.

DSF, 1991. Réseaux de surveillance de l’état sanitaire desforêts (réseau CEE et réseau Bleu): protocole pour les ob-servations. DERF, Min Agri For: 27.

Dahane B, 2006. Incidences des facteurs écologiques sur lesaccroissements annuels et la qualité du liège de quelquessuberaies du nord-ouest algérien. Mag Dept Forest FacSci, Univ Tlemcen. 129 pp.

Delatour C, 1983. Les dépérissements de chênes en Europe.Revue forestière française 35(4): 255-281.

González-Adrados JR, Montero G, Ortega C, 1993. Caracte-rización productiva de los alcornocales catalanes. InvestAgrar: Sist Recur For 2(1): 55-69.

González Adrados JR, García de Ceca JL, GonzálezHernández F, 2005. Evaluacion de anomalies y classi-ficion del corcho. Congreso internacional “Alcornocales,

Fabricas y commerciantes. Pasado, presente y futuro delnegocio corchero”. Palafrugell. 16 pp.

Gourinard Y, 1958. Recherche sur la géologie du littoraloranais. Alger, Service de la carte géologique de l’Algérie.Nouvelle série 6: 111.

Kaïd Slimane L, 2000. Etude de la relation sol-végétationdans la région nord des monts de Tlemcen (Algérie). Uni-versité de Tlemcen: 129.

Lanier L, Abbas M, Bensaada M, 1986. Rapport de missioneffectuée à M’Sila Alger, Inst Nat Rech For: 3.

Molinas M, Oliva M, 1990. El suro i les seves classes. L’Estoig(2):33-34.

Pereira H, 2007. Cork: biology, production and uses. Oxford,UK, Elsevier.

Pereira JS, Bugalho MN, Caldeira MC, 2008. From the Cork Oakto cork: a sustainable ecosystem. Santa Maria de Lamas.

SPCAN-DGB, 2007. Red de Seguimiento de Daños en losMontes, Red CE de Nivel I). Manual de campo. MR y M.Ministerio de Medio Ambiente. Madrid: 66.

146 B. Dahane et al. / Forest Systems (2013) 22(1), 138-146